GEOPIC, ONGC, Deharadun

Email: pathakr16@rediffmail.com

10th Biennial International Conference & Exposition

P 430

Common Reflection Angle Migration for Improved Imaging-A Case Study

R.Pathak*, Arjeesh Gupta, Surendra Kumar, R.K. Goel, C.M. Varadarajan

Summary

Kirchhoff migration has traditionally been the leading implementation for application of depth migration to seismic data.

There are many reasons for this, such as efficiency, ability to image steep and even overhanging dips, and flexibility. However,

the limitations of Kirchhoff migration are well known and its inability to image more than a single arrival is a major constraint.

Downward continuation algorithms, on the other hand, handle all arrivals but their inability to image steep dips is a severe

limitation. Instead, artifacts caused by the “swinging action” of the migration often obscure the real targets and it is very

difficult to distinguish artifacts from geology.

Today, there is an ever increasing demand for advanced velocity modeling and imaging techniques to provide an improved

knowledge of subsurface structures in geologically complex areas as well as more accurate and quantifiable description of

reservoir properties. Common Reflection Angle Migration is a new ray based seismic subsurface imaging technology that is

recently presented for generating high-resolution, amplitude preserved, and angle dependent reflectivity gathers in the local

angle domain. Such local angle domain common image gathers (CIG) can be obtained from a multi arrival, ray based Common

Reflection Angle Migration (CRAM) creating a uniform illumination at the subsurface image points from all directions. The

Common Reflection Angle gathers are ideal input for Amplitude versus Angle (AVA) and prestack inversion studies since they

are amplitude and phase preserved. .In this present study pre-stack depth migration has been done using CRAM technology

and comparisons are made with conventional Kirchhoff migration results.

Keywords: Local angle domain, Angle Gathers, Azimuth

Introduction

In order to overcome the possible kinematic and dynamic

artifacts on common image gathers generated by common

offset and common shot Kirchhoff depth migrations that

may adversely affect determination of accurate reservoir

properties, a reconstruction of common image angle

gathers are needed (Xu et al. (2001), Koren et al (2007).

CRAM (Koren et al, 2002 and Koren et al, 2008) is a multi-

arrival, ray-based migration that uses the whole wave field

within a controlled aperture. Unlike conventional ray-

based imaging methods working in depth-offset domain,

the ray tracing is performed from image points up to the

surface, forming a System for mapping the recorded

surface seismic data into the Local Angle Domain at the

image points. (Figure 2) CRAM's imaging process

combines a number of ray pairs representing the incident

and reflected/diffracted rays from the subsurface. The

procedure is based on a uniform illumination at the image

points from all directions, ensuring that all arrivals are

taken into account while amplitudes and phases are

preserved.

CRAM is specifically designed for detailed velocity model

determination; target-oriented, high-resolution reservoir

Imaging; accurate AVA and reservoir property extraction;

and imaging data recorded in areas of complex structure

and velocity. The migration supports isotropic and

anisotropic models, and can be performed using all types

of marine and land datasets, including OBC/OBS. The

CRAM algorithm is extremely versatile; thus it can be

adapted to any exploration objective. It can be used for full

volume imaging with full-apertures. It can also be run over

small target areas of interest with background dip azimuth

information, leading to a model-driven aperture for

achieving fast turnaround, high-quality and high-

resolution.

2

Figure 1: Location map of the area

Figure 2: Imaging in local angle domain.Picture demonstrate an

example of a selected ray pair incident and scattered at a given

subsurface point M and the four angles associated with the LAD

(local angle domain): dip v1 and azimuth v2 of the ray pair

normal, opening angle γ1 and opening azimuth γ 2. Four scalar

angles describing the directions of the incident and reflected rays.

Methodology

3D depth migration has created many opportunities for

exploration and production. At the same time, it has

brought many challenges. An essential prerequisite for

accurate imaging in depth is, of course, an accurate

estimation of velocity earth model.

For successful running of CRAM conditioned CMP

gathers and a précised interval velocity model is required.

It is recommended that gathers should be unregularised;

however it can be done in shot and receiver domain. Now

in updated version of CRAM provision of regularization

and interpolation has been invoked.

In this study conventionally de-noised and conditioned

DECON CMP gathers were taken as input. Stacking

Velocity analysis was done on these gathers at 1.0 * 1.0

km. Stacking velocity thus obtained was assumed as RMS

velocity, which in turn was used to run target oriented

Kirchhoff PSTM on every 20th Inline. The RMS velocity

analysis was done at 0.5 x 0.5 km on these PSTM gathers

and a refined RMS velocity volume was generated from

this velocity. KPSTM stacks were also generated for QC.

To move forward from this juncture the following essential

processing steps were executed:

Initial Depth-Interval Velocity Model Building

Initial Depth-Interval velocity model was created using

Constrained Velocity Inversion (CVI). This application

enables to create a smoothed and physically credible

velocity volume.

Target Line Kirchhoff Depth Migration for

Velocity Modeling

Kirchhoff Depth migration with Initial Depth-Interval

velocity model was run on target lines for estimating

residual move out and updation of velocity field. Suitable

Migration full aperture of 6000 m was used and output was

generated for every 20th inline for velocity refinement.

Automatic Residual Move out Picking

Automatic residual move out analysis was carried out on

target oriented KPSDM gathers and a residual volume was

generated, which is one of the major input being used in

grid tomography during the process of velocity updation.

Automatic residual move out correction is a procedure

which uses AVO technique in order to automatically

perform detailed residual move out analysis. The

application uses AVO technology and enables on the fly

enhancements of data prior to velocity analysis. Various

filters, amplitude corrections and mute options can be

applied for improved velocity analysis. The workflow

starts with the utility that allows selecting parameters and

viewing the results of the procedure on the selected

gathers. By optimizing parameters, we can obtain best

possible results in terms of the flatness of the gathers. Once

satisfied, the residual move out estimation can be

performed on the entire input volume. The output of the

procedure is residual move out volume.

3

Figure 3: Ray trace modeling for optimization of CRAM

parameters. Picture demonstrates the ray tracing from image

point to the surface taking interval velocity volume as

background. Rays in green color are successful rays which

received on the surface for a given offset, aperture and LAD

parameters.

Updation of Initial Interval Velocity using grid

tomography

3D grid tomography is a method for updating subsurface

seismic velocities. The model update is driven by a

measure of non-flatness of pre-stack migrated gathers. The

output of the 3D grid tomography is an updated subsurface

interval velocity model/volume. Since it properly

considers the contributions of different parts of the

subsurface in forming the image of an output trace, the

tomography has the potential to produce highly accurate

results. The tomography equations relate travel time errors

measured on the migrated gathers to the subsurface model

change, through rays which are traced from subsurface

reflectors to the surface. In grid tomography the updates

are calculated on a spatial grid which is generally coarser

than the velocity volume.

After the calculations are completed, the updates are

interpolated to the size of velocity parameter grids. The

interpolated updates are added to the initial volumes to

produce new velocity volume. 3D grid tomography is

divided into two parts which need to be run separately. The

first part constructs the tomography equations from the

input data, and in the second part, equations are solved. In

first pass of iteration KPSDM gathers, residual move out

and velocity volume were taken as input to this procedure,

which generated updated velocity volume. Thus final

updated volume was obtained after 4th iteration which was

used for running KPSDM and CRAM.

Depth migration using CRAM (common

reflection angle migration) technique

After being satisfied with the results obtained from running

final pre-stack depth migration and interval velocity model

CRAM (common reflection angle migration) was run. The

input used for the CRAM was taken as conditioned gathers

(CMP) and the final interval velocity model which is

obtained from the final iteration of tomography.

The parameters like opening and directional angle are

tested using the ray tracing that gives the number of failed

and successful ray counts. Final values are decided by

taking into account the more successful rays and by testing

those parameters on some inline having the complex

structure. The ray trace modeling is illustrated in figure 3.

The critical parameters for the testing are

1. Opening angle

2. Directional angle

3. Pencil distance

4. Azimuth

5. Aperture

6. Frequency

7. Reference depth for defining the angles at top and

bottom

The final parameters used in CRAM

Opening angle 60 (top) / 30 (bottom)

Directional angle 40 (top) / 40 (bottom)

Pencil distance 100 m

Aperture 6000 m x 6000 m;

Reference depth At top: 0 m ; at bottom:

6000 m

Frequency 60 Hz

Offset 6100 m

Fold 60

Table 1: Final parameters used in CRAM

Area of study

The area under study is located in MO block, Western

Offshore basin of India. Mumbai Offshore Basin is a

pericratonic rift basin, located on the central part of the

continental shelf of Western India and is the most

4

prominent of all the west coast offshore basins. (Figure

1)The basin is delimited by Deccan trap outcrops to its

north as well as east, west margin basement arch to its west

and Vengurla arch to its south. To its northwest, the

boundary between Bombay offshore and Saurashtra

offshore has been considered along the westward

extrapolation of Landsat lineament, which roughly

coincides with the limit of Neogene sediment exposures on

the southern part of Saurashtra Peninsula. Mumbai

Offshore Basin is an extensional passive margin basin,

with its genesis linked to the break-up of Madagascar from

the Indian plate in Cretaceous, about 88 ma. The basin is

contiguous with the petroliferous onland Cambay Basin,

through the Gulf of Cambay in the northeast.

The study pertains to the Pre Stack Depth Migration of 360

sq. km 3D seismic data, using Common Reflection Angle

Migration (CRAM) technology. The 60 fold 3D data was

acquired using six steamers with group interval 25 m and

shot interval 50 m during 2004 by an offshore vessel using

240 channels.The minimum offset is 100 m and the

maximum far offset is 6100 m .Record length of data is 6.0

sec. sampling interval 2.0 ms. The dimension of bin size

is12.5 x 25 m. The main objective of processing was to

improve subsurface imaging in depth domain using

CRAM technology in a zone of 1500-4500m depth.

Figure 4: Comparison of a Kirchhoff (a) and CRAM (b) PSDM

stack along in line L1.

Discussion

Here pre stack depth migration has been carried out using

Kirchhoff as well as CRAM imaging technique and results

are critically examined. In order to demonstrate the result

some representative sections are presented. Post

processing parameters for KPSDM and CRAM are same.

Figure 5: Comparison of a Kirchhoff (a) and CRAM (b) PSDM

stack along in line L2.

Figure 4&5 illustrate comparison between KPSDM and

CRAM images along in line direction. Looking around the

deeper part of the images it is found that KPSDM images

are defocused and blurred while CRAM images are highly

focused and sharp. This would certainly be helpful to

interpreter for understanding the subsurface geology in

complex areas.

Figure 6&7 show comparison between KPSDM stack

sections along cross line direction at two different

locations. It is clearly envisaged that over all continuity,

fault pattern definition, and relative amplitude preservation

are better in images generated through CRAM.

Figure 6: Comparison of a Kirchhoff (a) and CRAM (b) PSDM

stack along cross line C1.

In figure 8&9 depth slices generated through these two

imaging techniques have been compared at the same level.

It is observed that structural features are more clear and

sharp in the case of CRAM. This can give a lead to

interpreters in delineation of sub surface structures.

5

Figure 7: Comparison of a Kirchhoff (a) and CRAM (b) PSDM

stack along cross line C2.

It may also be noted that the CRAM gathers contain events

that appear less noisy and more continuous when

compared with the Kirchhoff offset gathers converted to

angle gathers. The higher angle coverage and improved

continuity suggests that the CRAM gathers would be more

suitable for use in AVA studies and other higher order

reservoir characterization workflows. (Figure 10)

Figure 8: Depth slice from the volume generated through

KPSDM.

Figure 9: Depth slice from the volume generated through

CRAM.

Figure 11 represent a CRAM gather at structurally

complex location of a section. It is seen that gather is flat

and noise free at locations where higher dip and undulation

is observed.

The well tie analysis was also done at different wells and

it was found to be in the range of 2 m to 72 m, well within

permissible limits. This is demonstrated in figure 12

showing an arbitrary line passing through wells.

Figure 10: Comparison between KPSDM offset gather converted

to Angle and CRAM gather. Gathers generated by CRAM are in

angle domain.

Unlike conventional ray-based imaging methods (e.g.,

Kirchhoff or beam migrations), the ray tracing in CRAM

imager is performed from the subsurface grid points up to

the surface. The procedure is based on a specially designed

point diffractor operator that ensures maximum

illumination of the image points from both, all subsurface

directions and all surface source-receiver locations, where

all arrivals are taken into account and amplitudes are

preserved in a better way.

Conclusions

CRAM is a very good tool for pre-stack depth

migration which generates angle gathers that are

optimal input for AVA/AVO analysis.

The final processed CRAM depth stack volume has

shown very good reflection character and better

resolution in the zone of interest.

CRAM images are more focused and crisps as

compared to KPSDM.

CRAM outputs show sharper imaging which

resulted in better fault definition. CRAM gathers

have appeared to be more clean and flat.

Deeper imaging is better and shows significant

variation as compared to processed output

generated through Kirchhoff PSDM.

6

Better preservation of relative amplitudes has been

claimed in CRAM outputs.

Frequency variation with depth appears to be more

realistic in CRAM outputs and results are expected

to help in mapping the gas chimney.

CRAM generates angle gathers which can directly

be used for AVA study.

Figure 11: CRAM stack section and gather in complex geological

situation .Gather is shown at a position P marked by blue vertical

line.

Figure 12: An arbitrary line passing through wells A, B, C and

D.

Acknowledgements

Authors wish to express their sincere gratitude to Director

(E), ONGC for giving permission to publish this paper. We

express our heartfelt thanks to Shri Anil Sood, ED-

HOIGEOPIC and Shri D.Sar, for their excellent guidance

and valuable suggestions during the execution of this

project. Sincere thanks are to Shri T.R.Murali Mohan,

GM- Head Processing, for his thoughtful supervision and

constant encouragement. Thanks are also due to the

interpreters from Western offshore basin, Mumbai India,

for their time to time interactions. Thanks are extended to

Heads of Software and Hardware Divisions along with

their team members and support services for lending time

to time support and resolving related issues.

Last but not the least; we are thankful to all the members

of the group, for their kind cooperation and valuable

guidance.

References

Dean K. Clark et al., Common Reflection Angle Migration

(CRAM) for improved input to reservoir description – An

example from, offshore India. Extended Abstracts ,SPG

India,Hyderabad,20010.

Koren, Z., Ravve, I., Barttana, A., and Kosloff, D., (2007),

Local angle domain in seismic imaging, In Expanded

Abstracts, EAGE, London.

Koren, Z., Ravve, I., E. Ragoza Barttana, A., and Kosloff,

D., (2008), Full azimuth angle domain imaging, 78th

Annual International Meeting, SEG, Expanded Abstracts,

2221-2225.

Koren, Z., and I. Ravve, 2010, Full-Azimuth Subsurface

Angle Domain Wavefield Decomposition and Imaging;

Part I: Directional and Reflection Image Gathers;

GEOPHYSICS, VOL. 76, NO. 1

(JANUARY- FEBRUARY),2011; P. S1-S13

Koren, Z., X. Sheng, and D. Kosloff, 2002, Target-oriented

common reflection angle migration: 72nd SEG Exposition

and Annual International Meeting, Expanded Abstracts,

1196-1199.

Ravve, I., and Z. Koren, 2010, Full-Azimuth Subsurface

Angle Domain Wavefield Decomposition and Imaging;

Part II: Local Angle Domain; GEOPHYSICS, VOL. 76,

NO. 2 (MARCH-APRIL 2011); P. S51-S64

Xu, S. H., Chauris, G.,Lambare, and M.

Noble,2001;Common angle migration :A strategy for

imaging complex media; Geophysics 66,pages 1877-1894.

Views expressed in this study are those of the authors only

and do not reflect the official views of the company i.e. the

Oil and Natural Gas Corporation Ltd.